Qwen3 30B A3B Instruct 2507
Qwen · Open weights · Released Jul 29, 2025 · qwen/qwen3-30b-a3b-instruct-2507
Best $0.048 in / $0.193 out via StreamLake · 128K ctx at this price (headline 262K)
262K contextToolsStructured outputOpen weights
Available from 5 providersUSD / MTOK · PRICES 23 MINUTES AGO · SPEED 2 DAYS AGO
EST. MONTHLY COSTclick a row for detail · click headers to sort
| PROVIDER | INPUT | OUTPUT | CACHE | CONTEXT | QUANT | TTFT | TPS | UPTIME |
|---|---|---|---|---|---|---|---|---|
| $0.048 | $0.193 | — | 128K | — | 2.4s | 192 tok/s | 99.9% | |
| DEDekaLLM | $0.09 | $0.30 | — | 262K | — | ✕ | ✕ | 99.8% |
| $0.09 | $0.30 | — | 262K | fp8 | 2.1s | 48 tok/s | 99.7% | |
| $0.10 | $0.30 | — | 262K | fp8 | 0.5s | 113 tok/s | 93.3% | |
| $0.13 | $0.52 | — | 131K | — | ✕ | ✕ | 100.0% |
TTFT/TPS: median (P50) over trailing 504h · chat10k profile (~10K in / 200 out) · measured via OpenRouter routing · ✕ = probed, endpoint returned no measurable stream · methodology v1.1 · measured 2 days ago
SERVE THIS MODEL AND NOT LISTED? LIST YOUR INFERENCE →
Recent changesRSS ↗
Oct 5, 2026endpoint removed · primeintellect6 → 5
Oct 4, 2026endpoint added · primeintellect5 → 6
Sep 12, 2026endpoint added · dekallm4 → 5
Sep 5, 2026endpoint removed · coreweave5 → 4
About
Qwen3-30B-A3B-Instruct-2507 is a 30.5B-parameter mixture-of-experts language model from Qwen, with 3.3B active parameters per inference. It operates in non-thinking mode and is designed for high-quality instruction following, multilingual understanding, and...
Specifications
Context window262,144
Max output32,000
Modalities intext
Modalities outtext
LicenseOpen weights
ReleasedJul 29, 2025
Knowledge cutoff2025-06-30
TokenizerQwen3
WeightsHugging Face ↗
Open weights
safetensors
Related
Embed badge
[](https://modelindex.ai/models/qwen/qwen3-30b-a3b-instruct-2507)Sources
OpenRouter API23 minutes ago
AGGREGATORHugging Face23 minutes ago
COMMUNITYModelIndex benchmarks20 days ago
MEASUREDVerified 23 minutes ago · list prices, not negotiated rates.